Motorcycle power system question

[Fake Name]

Hello!

I hail from the land of motorcycles, and I have a power generation issue regarding single-phase vs 3-phase system design. I'm well beyond my specific area of expertise, but need to choose a system wisely, as I travel on my bike off-road and well beyond civilization and prefer not to walk back.

Essentially, there is some controversy about specific design techniques of these systems, and I'd first like to know if I'm in the right place to ask. I promise I'm not fishing to develop something as a business or product for sale.

Is there someone here who might help me?

 

[berkeman]

Hello!

I hail from the land of motorcycles, and I have a power generation issue regarding single-phase vs 3-phase system design. I'm well beyond my specific area of expertise, but need to choose a system wisely, as I travel on my bike off-road and well beyond civilization and prefer not to walk back.

Essentially, there is some controversy about specific design techniques of these systems, and I'd first like to know if I'm in the right place to ask. I promise I'm not fishing to develop something as a business or product for sale.

Is there someone here who might help me?

Welcome to the PF.

Can you post links to some of the electrical generation systems that you are looking at? How much power are you wanting to make? Are you running Baja 500 style lights or something?

 

[Fake Name]

Hi, berkeman and thank you for your reply.

Yes, my electrical demands are greater than one's average dirtbike.

Yes, I can provide you with all levels of detail- I just want to be sure I'm in the right place first. I typically prefer to offer something to an online community before asking of it, but I've arrived with empty pockets, I'm afraid.

So- if I'm in the right place, I'll post tomorrow with details and photos of the components in question. In short, the original system is split ac/dc- ac for spark and lighting, dc for accessories and battery charging. The first stage of modification is to ac for spark ONLY, dc for accessory and battery charging. Then the next step of modification is the 3-phase system, typicaly found on streetbikes where the demand for power is dominant over the quest of lightweight. It's still AC for spark, DC for everything else, but provides a LOT more power. Stock is about 35 wattts, first mod is about 90-100, 3-phase is close to 200.

Thanks again for your reply- this sounds like something with which you can help me? I need someone outside the moto-community, as there are a lot of folks with products to sell and priorities other than science.

 

[Fake Name]

Uh-oh. Just noticed your avatar...you're not in the business, are you?

 

[pantaz]

Welcome! You will find plenty of two-wheel gear-heads around here. (Myself: XR400 dual-sport, FJ1200, RM400)

Simplistically speaking, 3-phase is more efficient.

 

[Fake Name]

OK, thank you in advance for your willing spirit, and welcoming my as a guest on your forum. I'll try to provide as much pertinent detail as I can, please ask if I've left out anything. I'm trying to sort through simple physics, opinion and internet lore. Some of this detail is too much- but I'm hoping I can be clear for the non-moto types as well.

First, the bike began as a 2003 KTM 525 MXC. It's a desert-racing, enduro dirtbike that someone has licensed for the street. KTMs are known as "ready to race" and are built for lightweight and horsepower. The bike has CDI ignition timed by a weight on the flywheel that sends a pulse to the cdi box via an ignitor coil. Pretty standard stuff.

The stock electrical stuff is a stator, which is stationary and the magnet-laden flywheel is attached to the crankshaft and encloses the stator.

The stock stator looks like this:

Notice it has 8 poles of varying quantities of wire in the coils. One pole appears quite different from it's 7 friends- as I understand it, that s the AC coil. In stock form that coil creates energy to power the cdi box (for spark) and the headlight circuit. That means in the event of a battery failure, the bike will still run enough to ride, but will not charge the battery, eventually resulting in the loss of electric start ability. But will still kickstart and run. The DC side is run through a regulator/Rectifier for DC battery charging and accessories.

OK so far?

This system puts out an estimated wimpy 35 watts. It's just enough to do the job without adding weight and expense, and is simple enough to offer uncomplicated reliability.

Along comes the crowd, myself included, for whom nothing is good out of the box. We have to- COMPELLED to modify. One modification is a headlight that exceeds a single candle on a windy day. My personal choice is the Trail Tech Extreme Race Light, 95 watts of HID . It's a LOT of light. I'm old, my eyes are old, and I can't see a damn thing anymore. And it's cool- here, look:

I also add heated handgrips, two GPS (one including XM) and possibly a heated vest. Demand clearly outstrips supply.

The most popular modification is the TrailTech stator and DC modification. This involves replacing the stock stator and adding a new regulator/rectifier, then converting all non-ignition circuits to DC.- It's an easy modification, and here's a snappy of the stator TrailTech sells with it's friend the regulator:

Notice the 8 pole design, larger gauge wires filled further than the stock stator. Again, one pole visually different, which I assume is the AC pole to power the cdi box.

Trailtech makes it easy- almost plug-and-play. Dead reliable, this is the standard modification- done by thousands of riders worldwide.

But it's not enough power, Scotty. TrailTech claims 100 watts, but I suspect it's something shy of that. If I run all three lamps of my bitchin headlight, I'll discharge my battery. I have a digital voltmeter on the bike to monitor such things- voltage drops to around 11.2 to 11.7 at speed, at idle speeds insufficient to even keep the lights on.

It breaks my heart.

So- along comes an alternative. A few years ago, a company called ElectroSport began creating 3-phase systems for dirtbikes. A 3-phase system is most commonly found on street bikes, and they have a greater power demand, and care less about weight. Electroposrt began selling it's products and just as quickly began to have serious problems. Fried and boiling batteries, non-start problems, including one that would cause the bike to fail estarts, but work fine with the kickstarter (more on that later).

Electrosport claimed a manufacturer had substituted components of a lower quality without their approval, and this caused failures in regulation, etc. It ended up being a nightmare. Finally, one EE was hired to help out, and he did, but t was too late. The reputation of the company was shredded, and it was eventualy sold in shame.

Fast forward a few years. The EE that was brought into help Electrosport has started his own company creating 3-phase power systems. As I lust for power, I decided to try them out. EE has assured this system is similar to, but improved from the Electrosport system. I'm a commercial photographer by trade, and I like to help fledgling businesses get going, so I offered my services.

Here is what the new system looks like (beautiful photos no?):

and the regulator/rectifier- clearly far more robust with greater heat-dissipating ability:

I installed this setup on the bike and fired it up. PLENTY of power (Scotty)- this thing must have Dilithium crystals in it. With all demand on the system, it still provides 14.5-14.8 volts at speed, and just over 12v at idle. Perfect, right?

Well, maybe. My first long ride with the system failed at mile 113, between Palmdale and Barstow. Subsequent troubleshooting at home led to no-spark issue, which led to a failed AC pole. EE claims it a manufacturing defect, the first he's had in over 500 units.

OK, it happens.

But here's where it gets more interesting, and more internet-based chatter. And, frankly the reason for this winded post.

Notice the 12-pole design? There's another player online who makes highly custom and very expensive 3-phase systems. He eliminates the AC pole completely, and runs the entire system on DC. His claim is that the design of the stator- it's 11 DC poles- is inherently unstable and "unbalanced". He claims a system like this can never be relaible and is based on flawed science.

Do you agree with this analysis? Is the number of poles a factor when creating a reliable power system in three phase?

Further, notice the AC pole is smaller than the previous versions? This stator will start the bike using the electric start MOST of the time. If the battery is at it's peak voltage, and the fuel/air mixture is just right, it fires up like a champ. But if the bike needs to be cranked a bit more, dropping battery voltage, I have to resort to the kickstarter, which will turn the motor faster thus providing more AC power to the cdi box.

Do you believe the AC pole is too small to create enough AC for reliable ignition at lower (estart) rpm?

Could another pole be used to create more AC- use two poles instead of one? And (bonus points) could that be wired in a way to provide redundancy in the case of a (another) single-pole failure?

Whew.

Here's a pretty picture of me actually riding:

Thank you in advance for your thoughts, and please post questions- I'll do my best to answer.

 

[pantaz]

...

What you're calling the "AC pole" is the ignition pulser coil. You can test it with an ohm meter (disconnected from the wiring harness). Check you KTM manual for specs.

The unbalanced three-phase does seem unusual. Measure the ohms for each phase (each yellow wire is one phase). Connect the ohm meter between a yellow wire and the ground. (If there isn't a ground wire, use the metal body of the stator). All three phases should be very nearly the same. If one phase is different, then he might have done something in the electronics to compensate. I'll have to check some things about rectifying 3-phase before I can offer anything useful.

BTW: I don't have anything too recent, but I have photos online from riding Dove Springs/Jawbone, Las Padres, and a few other spots.
http://smg.photobucket.com/albums/v11/pantaz/LosPadresNF/

Visiting the Husky Monument (I'm on the left):

 

[Fake Name]

Thanks for the reply.

Please correct me if I'm wrong about this- my understanding is the ignition is timed via the electrical device ( I believe the technical term is a "Hall Sensor") that is bracketed immediately adjacent to the flywheel- completely independent of the stator, that senses a mass of metal attached to the outside of the flywheel. See number 5 in the diagram. Remember, the stator is static, and timing is (I think!) dependent on a dynamic input that will vary with the speed of the motor. So I don't see how that pole could be used to time the ignition. But I'm often wrong- just ask my wife.

The nomimal resistance for a acceptable stator coil is right around 12-15 ohms. The stator that failed checked in at 14, but the bike failed to run until it was replaced.

Is that correct?

Thanks again.

 

[Fake Name]

What you're calling the "AC pole" is the ignition pulser coil. If one phase is different, then he might have done something in the electronics to compensate. I'll have to check some things about rectifying 3-phase before I can offer anything useful.

I suspect not. This regulator is used across for many different bikes and stators in his product line- I know it's not stator-specific.

 

[pantaz]

... See number 5 in the diagram. ...

I used the wrong name. The coil on the stator portion is the exciter or trigger coil. It's not timing the ignition, but it is part of the CDI circuit.

The nomimal resistance for a acceptable stator coil is right around 12-15 ohms. The stator that failed checked in at 14, but the bike failed to run until it was replaced.

Regarding the starting issue, the power output of the stator isn't the issue, it's the trigger coil. You said it's physically smaller than the original version. If the new trigger coil design is supplying lower than OEM voltage to the CDI, then you'll have starting problems.
Replacing the stator unit also gives you a new trigger coil, which likely why your starting problem went away.

As for unbalanced three-phase... ugh. Almost everything I found is related to unbalanced loads. There is some data on three-phase rectifiers being damaged by unbalanced phase angles, which may have some relation to your unit's failure, but I can't say with any authority.

 

[Fake Name]

Ok- help me out, because I'm easily confused.

It's my understanding (easily wrong) the single pole on the stator creates energy to power the CDI box. That power is then sent via the CDI to the ignition coil to be transformed to a pulse voltage sufficient to create spark at the sparkplug.

The specific TIMING of that pulse is controlled by the message sent by the pulsar coil (#5 on the parts diagram)- a device separate from the stator- that communicates to the CDI box the specific position of the crankshaft.

Do I have that correct? I'm getting tripped up my the semantics- a trigger coil sounds like a timing device, and an exciter coil sounds like an ignition coil.

Here's a diagram, if that will help:

I "get" that the physical size of the newer stator's coil (ignition/trigger/exciter) would suggest an insufficiency of power at low RPM- granted we don't know how much the CDI box really needs, we just know the older stator provided more.

As I seem to have an abundance of power for battery charging and accessories, could one of the other poles be sacrificed and use TWO poles for ignition energy and the remaining for of the rest of the bike? And could those then-two poles be wired in parallel to provide redundancy?

Thanks again...

 

[jim hardy]

wow - fascinating thread...


my thought is

no pictures of overheated wires on your stator...i'll assume none were shorted through to the metal frame.

if it's like outboard motors
the "ac " pole is dedicated to charging the capacitors for your CDI. it has more turns of smaller wire (probably makes a few hundred volts)
and the others to making DC for the accessories
an even # of poles is common but i see no reason why eleven wouldn't do. after all , lawnmowers get by with one pole and you can't get more unbalanced than that...

the regulator works by shorting out the individual phases at point in the cycle to make average voltage at battery what it wants, ~14.3 or so

i'm wondering if the permanent magnets in your rotor have lost strength,,,,... in old days it was important to place a thin iron "keeper" over magneto magnets when disassembling.
the test for that is to measure the AC voltage produced by stator when regulator is disconnected , ie a 'no load' test on stator.
The old Suzuki i had wanted 40 volts AC at 2000 rpm i think - does your manual describe such a test?

if you find the magnets weak, a good magneto shop might be able to recharge them.

 

[Fake Name]

Thanks for joining in!

A typical stator- the "other kind" will put out an unregulated 7-28 ac volts. I've not measured the 3-phase stator.

Just got back from a 150 mile ride. Plenty of DC power- with headlight and grips on, it MIGHT hit 11.9v at idle, but anything above idle hits 14 easily.

But the bike will consistently fail to e-start, but fires right up on the kickstarter.

So- back to one of the original questions- nothing about this stator design (the section dedicated to DC) suggests a circuit umbalance? No clipped waveforms that might cause regulator overheating? No weird voltage spikes or anything like that?

Not hearing anyone have any design issues, except the obvious ignition charging pole...

 

[jim hardy]

i see no design issue, from the information posted.

the system will be sensitive to cranking speed. that's because voltage it delivers to CDI is in proportion to rate of change of magnetic flux from rotor,
i.e. the product of magnet strength and rotor speed at instant magnet crosses that particular pole.

perhaps that explains your observed sensitivity to battery state-of-charge, manifesting itself as cranking speed.

easiest design tweak would be a few more turns on the CDI pole to give more volts.
but i'd measure the volts produced by your CDI winding first.
I did that on my outboard just to learn how it works...
On my outboard motor the "kill" switch went directly to the CDI energy storage capacitor , so the voltage across kill switch was my CDI winding voltage after being rectified into DC.
It was around 260 volts DC at slow idle, 100 volts cranking, 400 volts running.
(Obviously the system was very sensitive to dirt and moisture in vicinity of kill switch..)
Evinrude strategically placed that winding where you'd have good flywheel velocity when the magnet went by it - important when doing a rope start on a big outboard engine.

i suggest you make some measurements of stator voltage delivered to CDI both at idle and when cranking..
I notice there's no CDI module connection to battery on that wiring diagram so energy all must come from that CDI winding.
Condition of energy storage capacitor inside the CDI will be important, ,,
hopefully there's a place to measure the rectified DC across it (like there was on my outboard motor)

old jim

edit - don't ignore that trigger sensor. if it has a coil it too will be sensitive to speed, and to air gap. How do you adjust its closeness to rotor?

 

[jim hardy]

and it could be as simple as new brushes in starter...

 

[Fake Name]

Fresh (new, rebuilt) starter, but thanks.

So.

Sounds like the number and nature of the poles of a stator are not a factor in it's ability to generate a proper and complete waveform for power generation?

If that's true, then perhaps the stator body could be redesigned. If the "ignition pole" remains the proper height from the axis, and the proper depth (front to back in a side view) then that pole could be made wider by narrowing the other poles a little- say 10%. It's not a rotating mass, so additional weight (if any) is not a problem, right? And given I have plenty of DC power, I might be able to sacrifice some from those 11 to increase the size of the one.

So, by making that ignition pole wider, we'd have more wire in the windings- but is that enough? Does it need more turns in the coil to make more energy? Does the resulting new diameter of the pole change the output- is a larger area sufficient to make more energy? Seems like it to me, but heck, I'm a photographer, so what do I know?

I realize we can't quantify the actual results without testing first the output of the current pole against the output of our theoretical new dimensioned pole.

But the theory stands, correct?

Thanks again, everybody.

 

[Fake Name]

PS: The pulsar coil is set a fixed distance (uses a bracket that is not adjustable) from the flywheel, and the coil itself is fairly new- less than 200 hours.

 

[jim hardy]

your bigger pole will have to retain alignment with the rotating magnets. in theory giving it a little more area will increase flux hence output. how far can you push it is a question for the magnetics guy who selected the parts.
my intuition says to try a few more turns on the existing iron.
another question for your magnetics guy is - can we zap the rotor a bit harder and get more flux, for sake of easy starting.

http://www.magnetapplications.com/assemblies.htm

keep us posted

btw - i really admire your photos. great depth of field, everything is visible! that would be a fun project, good luck!

old jim (79 Sportster, looking for twin Enfield to complement)

 

[pantaz]

... If that's true, then perhaps the stator body could be redesigned. ... So, by making that ignition pole wider, we'd have more wire in the windings- but is that enough? Does it need more turns in the coil to make more energy? ...

Yes, more turns = higher voltage. I can't answer, "is that enough", since we haven't yet identified the specific deficiency.

The exciter coil uses very fine wire -- 35 ga. (0.0056 inch) on the one I rewound. You could add quite a few more layers before the size would interfere with anything.

 

[Fake Name]

Which raises another question (for me anyway)-

Is it the number of coils, or the length of the wire in the coil?

If the pole is just as "tall" and "deep" and we just increase the "width", that would not really increase the number of trips around the pole that the wire can make. You'd end up using more wire, yes- but as each coil is larger there might not be a big difference in number of coils.

Or am I wrong?

 

[jim hardy]

volts induced is the product ::

number of turns X (rate of change of) magnetic flux encircled by those turns

since at idle your flux is swapping its direction back and forth between whatever is strength of magnet,,
you get so many volts per turn... and as you speed up the flux swaps direction more often so rate goes up and so does voltage.
probably that's why a swift kick starts it reliably (assuming good rings&valves)

a stronger magnet should give more flux at idle hence more volts
a larger pole (more area) or a smaller air gap (less clearance) should allow the magnet to push more flux , hence more voltage;
more turns should allow the same flux to produce more volts...

obviously more turns is easiest for the home hobbyist, no machine work required

but one experiment is worth a thousand expert opinions.
it would be interesting to try a few battery starts with voltage regulator unplugged just to rule out an unsuspected interaction between charging and CDI poles...

there will come a point where additional turns don't give more volts because of something called "Armature Reaction" - wherein the load current in your coil significantly opposes the permanent magnet. your magnetic design guy knows how close he has pushed that.

e = nXdphi/dt
e=volts, n= number of turns, dphi/dt=rate of change of magnetic flux ...
you can increase flux by increasing intensity (stronger magnet), by increasing area(makes more room for flux) or by decreasing airgap(makes it easier for magnet to push it through)...

or you can increase n, the number of turns that encircle whatever flux you have.

hope this helps you paint a mental picture that works...
old jim

 

[Fake Name]

volts induced is the product ::


e = nXdphi/dt
e=volts, n= number of turns, dphi/dt=rate of change of magnetic flux ...

old jim

Well, like...duh.

I knew THAT.

 

[Fake Name]

OK, I knew nothing of the sort.

And that's why I'm here instead of Facebook ; )

The designer/builder of these units is (at least theoretically) interested in producing a product that works. I'm interested in HIM producing a product that works.

He's currently working on a newer version for which I'll Guinea Pig, but the whole question has me (for some reason) intrigued.

The flywheel is a fixed asset- either KTM or one of the aftermarket that CLAIM higher-energy magnets, but yielded little, if any, difference for me in practical terms.

I suspect it's easier to redesign the stator poles than the flywheel.

 

[jim hardy]

hmmm just musing now

since pulsar is in a fixed location wrt TDC

that CDI box makes a decision about spark advance/retard based on flywheel speed, and probably throttle position too
and its thought process is a complete unknown at this point

i would just have to make myself an experimenter's bracket for pulsar coil
and see if changing its location by ten degrees or so affected start
and if so, discuss that with CDI circuit designer guy

lastly - is the orientation of CDI pole wrt TDC imnportant? does that stator mount in two different positions? I noted symmetrical bolt holes in your excellent photos...

 

[Fake Name]

CDI Circuit design guy? Probably in Austria or Japan...

The bracket on the pulsar coil gets that pickup VERY close to the flywheel. 10 degrees would make it a lathe.

Correct, the breathtaking photographs show the symmetrical mounting holes on the stator body, so it MIGHT be possible to mount the stator 180° from it's standard mounting position- but...

The cables that exit the back side of the stator fit into a groove machined into the stator cover casting, then clamped in position with a little bracket and a screw.

About the photos- after 30 years in the business, I'd sure the hell better make nice ones.

Here, have a look at the current and legacy site.

 

[Fake Name]

volts induced is the product ::

there will come a point where additional turns don't give more volts because of something called "Armature Reaction" - wherein the load current in your coil significantly opposes the permanent magnet.

Hmm. Will it reach a threshold of diminished returns, or will increasing the coil area begin to REDUCE the voltage yield?

 

[jim hardy]

Hmm. Will it reach a threshold of diminished returns, or will increasing the coil area begin to REDUCE the voltage yield?

diminishing returns... it's as much a natural law as gravity.
open circuit yield will always go up with more turns
but as you add more turns you'll bump against a limit on how much current you can draw

that's how the voltage regulator worked on my outboards and old Suzuki, it shorts out the coil
and enough current flows to cancel out the magnet but not enough to melt the wire.

taken to its logical extreme you could put so many turns on the pole it'd limit current to less than you need.
but open circuit voltage would be quite high.

have you access to an old stator, flywheel and lathe? one could really experiment then.